Exploring Connectivity: Southeastern Bengal and Beyond (c.

500-
1300CE)

Prologue

The region which is commonly called Bengal was in fact in early times an agglomeration of four sub-
regions,which was largely determined by its hydrography, (Bhattacharyya, 1977) of which the large
spatial segment to the east of Meghna was divided into two kingdoms known as Samataṭa and Harikela
in historical sources in different periods of time. Taken together they formed the Southeastern region of
present Bangladesh. The presence of Samataṭa in historical sources is earlier than that of Harikela. The
relevant sources in the form of pra śasti, copper- plate charters, coins, indigenous literature, Chinese
travel accounts and others indicate that it is not always possible to have any water tight demarcating
line for these two geo-political units. Though these two units had their core territories, like the Comilla-
Noakhali plain for Samataṭa and the Chiittagong coastal tract for Harikela, in different periods of time
their contours overlapped and sometimes one territory subsumed the other (Ghosh&Pal,2006:78-96)).
Thus the Trans Meghna region represents the present territories of Noakhali, Comilla and Chattagram
in Bangladesh and Tripura in India and embraced the Samataṭa and Harikela kingdom. Morrison with
his analytical study of the copper-plate charters of early Bengal (Morrison, 1980:115) was the pioneer
in indicating that the Samataṭa region was culturally different from other areas of the delta and further
researches supplement his position. In fact the geographical location of the region helped in evolving
its own cultural ethos with a pronounced sub-regional identity. (Ghosh, 2010:220-231)

It is important to trace briefly how Samataṭa became an independent kingdom from being located in
the periphery of the Gupta empire. The term Samataṭa appears in the fourth century CE in the
Allahabad pillar inscription of Samudrāgupta as a pratyanta rājya (frontier state) together with
Davāka (valley of the Kapili-Yamuna, Kolang rivers) and Kāmarūpa (Guwahati region of
Assam), thereby indicating its proximity to Assam. It is referred to as frontier as the Gupta
stronghold was in the Ganga valley. The rulers of these areas are referred to as simply nṛpatis of a
particular region without even having their names mentioned. It is a pointer to the fact that these rulers
were perhaps merely chiefs and so did not gain enough attention from the Gupta ruler to be referred to
by name. During the time of the other Gupta rulers succeeding Samudrāgupta, we have no information
regarding Samataṭa. That kingdoms flourished in the Comilla and Faridpur areas in the 5 th and 6th
centuries is however attested by epigraphic records.
[[ The Gunaighar copper plate dated 507 CE suggests that Mahārāja Vainyagupta was ruling in the
region. The copper-plate mentions a division called Uttara maṇḍala and a jayaskandhāvāra at Kripura
which are likely to have embraced a small and compact area within the broader region of Samataṭa.]]

Meghna, which practically separates Samataṭa from the rest of Bengal, formed the western boundary
of Samataṭa and this explains its separation, from Vaṅga as a geographical entity as well as its
association with Davāka and Kāmarūpa as frontier belt of Samudrāgupta’s empire (Bhattacharyya,
1977:67).
Chinese travellers like Xuan Zang (c. early 7th century CE) and Sheng-chi (2nd half of the 7th century
CE) also mention Samataṭa in their records. In the period following the disintegration of the Gupta
empire, we have numismatic evidence to prove Śaśāṅka’s rule in the Samataṭa area. A very specific
type of gold coin with higher gold content than coins in other areas of his domain was issued by
Śaśāṅka in Samataṭa. B.N.Mukherjee rightly argues that the issuance of better quality gold coin by
Śaśāṅka was deliberate keeping in mind the trading potentiality of the region.. Thus Śaśāṅka must have
had some control over Samataṭa and so he ensured the circulation of good quality gold coins.
In the second half of the seventh century three local powers, known as the Rātas, Nāthas and
Khaḍgas were ruling Samataṭa and its environs. In the seventh century CE, Samataṭa emerges as an
independent monarchical set up, one of its kings claiming the designation of Samataṭeśvara or Lord of
Samataṭa. [[The Kailan plate of Sridharana Rāta (c.665-675) shows that he and his father
Jivadhārana bore the epithet of Samataṭeśvara and had the hill fort of Devaparvata encircled by the
river Kṣirodā was their administrative headquarter.]]

The Chinese priest Sheng-chi also found Rājabhaṭa, the Rāta ruler ruling over Samataṭa which
indicates that Samataṭa was coterminous with the kingdom of the Rātas. The Khaḍgas(c. 625-705 CE)
on the other hand had Karmānta (Barkanta in Comilla) as their royal residence wherefrom they issued
their early land grants (viz., the Ashrafpur plates). D.C.Sircar suggests that they originally ruled in
Vaṅga (Dhaka-Faridpur-Bakharganj area and adjoining regions) and Devakhaḍga (c.658-73 CE)
extended his power to Samataṭa. It may be pointed out that both of these two lines of rulers owed their
allegiance to a superior power. Towards the end of the seventh century or in early eighth century, the
Khaḍgas were supplanted by a Deva family. Ānandadeva, a ruler of the Deva family calls himself
Vaṅgālamṛgān͘ka and had his new capital at Vasantapura (year 39) whereas the old capital was
damaged by an invasion of a strong enemy force and had to be abandoned. The question that needs to
be addressed here is why Ānandadeva is using the epithet ‘Vaṅgāla’ as we know that Vaṅgāla was a
separate sub-region and we have no direct evidence of Deva rule in Vaṅgāla. Vaṅgāla adjacent to
Vaṅga (Dhaka,Vikrampur & Faridpur) of the early medieval period is the coastal region of present day
south-eastern Bangladesh. According to D.C.Sircar, probably by Vaṅgāla, the Bakharganj subdivision
was meant and may be the Devas were ruling in this area. The rule of the Devas in Samataṭa perhaps
suffered a blow with the occupation of the region by Gopāla (c.CE 750-775), the founder of the Pāla
dynasty, as indicated by the Sian inscription of Nayapāla. It is true that this inscription refers to the
association of the Pālas with Samataṭa but it was perhaps not the earliest political stronghold of the
Pālas as suggested recently . We have no other evidence except the Sian inscription of a later period
and the image inscriptions of the time of Mahipāla I (c.977-1027 CE) to suggest the presence of the
Pālas in Samataṭa. On the other hand we know of a ruler called Bhavadeva (c.CE775-800),
contemporary to Dharmapāla, issuing copper plate from Devaparvata, the capital of Samataṭa.
Moreover if Samataṭa would have been an early political stronghold, the records of the early Pāla rulers
would have referred to the possession of Samataṭa or issued copperplate charters from the area. It
appears that the local Deva dynasty recovered the region after the initial occupation of Samataṭa
by Gopāla. Mahipāla I could have reoccupied the Samataṭa area defeating the Chandras, who had by
then integrated Samataṭa within their kingdom. Abdul Momin Chaudhury also opined that sources do
not suggest that Gopāla rose to power in southeastern Bengal. Though Rāmacharitam mentions that
the royal lineage of the Pālas originated from the ocean, here we cannot take it as Samataṭa because
from the Sian inscription it is clear that the ancestor of Dharmapāla attacked Samataṭa, it was not their
place of origin. After this period till the time of the Chandra ruler Trailokya Chandra (c. CE 905-925),
the history of Samataṭa is unknown to us. We only learn that Trailokyachandra defeated the Kambojas
in Samataṭa.We can conclude that from the 7th century to the 9th century,Samatata was
considered an integral, independent political stronghold.
In the middle of the 9th century Samataṭa’s status of an independent kingdom suffered a blow
with the Chandra occupation of Samataṭa. [[According to the Paschimbhag copper plate his army
conquered Samataṭa and captured its capital Devaparvata and attacked Vaṅga. ]]
Shortly before the Chandra occupation of Samataṭa, the city of Devaparvata appears to have been
devastated by the Kambojas. . One may however note that with the coming of the Chandras, there is a
shift in the location of politico-administrative centre from Devaparvata to Vikramapura. This was due
to the more extensive power base of the Chandras which necessitated a shift in the capital and to a
more viable geographical location. Thus during the reign of Chandras Vikramapura loomed large as an
important administrative centre and Samataṭa was transformed to a mere administrative division like
maṇḍala. Its independent political identity lost. Loss of Samataṭa’s political identity however led to the
integration of the two kingdoms under a single political authority.

The coherent political character of the sub-region was disturbed by the ouster of the Chandras in
the second half of the 11th century by the Varmans and finally we find that during the later part of
the Sena rule an independent dynasty known as the Deva dynasty was formed in the eastern side
of the Meghna river. in the thirteenth century, Pattikeara (Patikara or Paitkara is a pargana near
Mainamati), which was within Samataṭa, formed a separate kingdom. Pattikera developed
relationship with Pagan. The Pagan-Pattikera relation has been highlighted by scholars drawing
examples from the matrimonial alliances between the two kingdoms. The marriage between King
Kyanzittha and Queen Abeyadana of Bengal provides evidence for Mahayana influence in Pagan and
later religious development. Two kings, Harikāladeva (1204-30CE) and Vīradharadeva (1230-
50CE) are known (no connection can be established between these two kings) and their records come
from Mainamati.

The name Harikāla is interesting. It may be an example of a territorial name being converted into the
name of a ruler. The second plate refers to the grant of land at Vataganga viṣaya (not identified) in
Samataṭa maṇḍala of Puṇḍrabhukti. Devaparvata lost its importance by this period and is not heard of
any more. Pattikera developed relationship with Arakan. From the Mainamati copper plate dated
AD1220 and issued in the 17th regnal year of king Ranavaṅ͘kāmalla Śri-Harikāladeva, it is learnt that
Dhaḍi-eba, a minister of the king, donated a piece of land in favour of a vihāra dedicated to Durgottārā
(a form of the Buddhist goddess Tārā) in the town of Pattikera. Dhaḍi-eba is noted as shining in the
good practices of the Sahaja cult. From his name and also the name of his father Heḍi-eba it has been
surmised by some scholars that the family might have been of Burmese origin (Bhattacharyya,
1933:282-289). Thus Pattikera as a kingdom developed at the very heart of Samataṭa which came to
form a maṇḍala under the Chandras and shrank in size. Gradually it lost its independent identity and
merged in to the expanding Paundrabhukti (which originally embraced north Bengal during the Gupta
rule). Dāmodaradeva (1230-55AD) and his predecessors probably ruled as feudatories of Senas of
Vikramapura although Dāmodara’s allegiance was nominal. In his Mehar copper plate Samataṭa
appears as a maṇḍala . Available evidence till date suggests that the name Samataṭa can be last seen in
the Mehar copper-plate (Saka 1156 i.e. 1234 CE) of Dāmodaradeva (Sircar, 1982:47).

It will be in order now to look at the history of the other geo political space, the immediate eastern
neighbour of Samataṭa, i.e. Harikela, which appeared in inscriptional, numismatic as well as textual
sources from about the seventh / eighth century CE. According to I-jing (travels in India 675-695CE)
(Takakusu, 1896:44) who met Wu-hing near Nālanda in the first year of Ch’ui-king period (c.685 CE),
the latter sailed from Simhala for the northeast and “came to Harikela, which was the eastern limit of
Eastern India [Tung T’ien (-chu)] and of Jambudvipa”, (Mukherjee, 1975:115-119). This has been
taken to suggest that Harikela must have had a littoral area within its limits. In the Glossary of Hui-lin
(817 CE) Samataṭa, Ho-lai-ka-lo and Tāmralipta were placed near Kāmarūpa (Bhattacharyya,
1977:70). It is not clear whether it was an inland territory or a littoral area or both. In indigenous
sources Harikela finds mention at about the same period. In the Ārya Mañjuśrī Mulakalpa (8th century
CE), (Sircar, 1982:48), Vaṅga , Harikela and Samataṭa are cited as distinct entities. I-jing also refers to
Samataṭa thereby indicating that Harikela was that region which lay to the east or south-east of the
Comilla and Noakhali districts (Takakusu, 1896:44). He was also aware of sea-borne journeys from
Java and the Malay Peninsula to Holaikalo or Harikela . According to him Tan –Kwong, a Chinese
priest came to India by the southern sea route and having arrived at A-li-ki-lo (Harikela), he was
reported to have found much favour with the king of Harikela.(Beal, 1973:XXXIX) This can perhaps
be taken as the earliest indications of Harikela’s maritime network in the Indian Ocean. Thus the name
‘Harikela’ denoted by the second half of the seventh century CE only the coastal area of Chittagong .

It appears that a Buddhist line of rulers ruled in Chittagong. To it belonged Devātideva, Kāntideva and
Attākaradeva although no connection can be established between them (Bhattacharya, 2000:471-487).
A land grant of Devātideva dated 715 C.E. refers to Harikelayam (the people of Harikela). Till date this
is the first epigraphic reference to Harikela. The Chittagong area over which he ruled is described as
Khasa-maka. Probably Devātideva belonged to the Non-Aryan Khasa tribe (Bhattacharya, 2000:471-
487) mentioned as mercenaries in the land grants of the Pālas. The Kshatriya character of the Khasa
people which they acquired through their military prowess are highlighted in a number of literature.
(Ghosh,2006: 206). This Kshatriya character of the Khasa people is clearly pronounced in the metal
vase inscription(fig.1) where the ruler Devātideva belonged to the Khasa tribe. Though we have no
information about his origin in the vase inscription, we may suggest that the Khasa tribe resided at the
Chittagong hill tracts. It is specifically mentioned in the inscription that land transaction was made
from the residents of the Chandrabhaṭṭārika grāma situated in the Khasa maka (Khasa kingdom).
Under Kāntideva (c.800-825) Harikela came to form a maṇḍala with its capital (vāsaka) at
Vardhamānapura identified with present Bara-Uthan village of the Patiya Upazila in Chittagong. If
Harikela was a maṇḍala, what was the name of the kingdom of Kāntideva is not known. Harikela
continued as a maṇḍala under Rājādhirāja Samaramṛgān͘ka Attākaradeva (early tenth century CE).
According to Gouriswar Bhattacharya, he was an Arakanese who ruled over this region. He was
probably a subordinate of TrailokyaChandra (c. 905-925), the father of Śrīchandra (Bhattacharya,
1993:323-338). The first epigraphic reference to Chattagrama was perhaps found in this vase grant of
Attakāradeva datable to 10th century CE.(fig2.) It refers to an area as navachatta. The word nava
affixed to Chatta may suggest that it was a new settlement.(Harikelamaṇḍalantargata
navachaṭṭalasambaddha) Recently a coin showing the name Attākara has been found which is in line
with the Ākara coins.(fig.3) (Rhodes, 2006:76; Islam, 2014:163). If they are the same person, then he
belonged to the Ākara family of Chittagong. His capital was also Vardhamānapura. According to the
Rāmpāl plate and Bogra plate (Flemming,2010:223-244 ) of Śrīchandra, TrailokyaChandra (c. 900-
925) became king at Chandradvipa (Bakharganj) and he was the ‘mainstay’ of the royal family of
Harikela (harikelarājakakudacchatrasmitānāmśriyā[m).We have no idea about the overlords of the
Chandras. The statement has been taken to mean that this family rose to supreme power from the
position of a subordinate status in Harikela (Sircar, 1982:106). Later Vaṅga , Samataṭa and Śrihaṭṭa
formed the Chandra kingdom. It has been suggested that with this extension of the Chandra power to
Śrihaṭṭa from their base at Harikela the name came to be associated with the region of Vaṅga and
Śrihaṭṭa (Mukherjee, 1975:115-119). However it is to be remembered that the records of the Chandras
do not identify Vaṅga with Harikela or refer to Śrihaṭṭa and Vaṅga within the limits of Harikela . It is
only in the texts of Yādavaprakāśa (11th century) and HemaChandra (11-12th century), we have the
phrase Vaṅgāstu-Harikeliya identifying Harikela with Vaṅga .

The Hudud al-A’lam refers to H.rk.nd (taken to be the same as Harikela) as a place on the sea coast
together with N.Myas, Urshin, S.M.ND.R and Andras (Minorsky, 1937:87). The reference to Harikela
along with Samandar as places on the seacoast may be noted. Either it is a misinterpretation on the part
of the author as within the Harikela country lay the port of Samandar or it might so happen that
Harikela was the name of a kingdom as well as the name of a place identifiable with present day
Chittagong.

The most important evidence regarding Harikela is the sustained minting of silver coins bearing the
name Harikela (fig.4) These coins have been found from Mainamati, Jobra in Chittagong, Sylhet,
Belonia subdivision of Tripura , Sandoway in southern Arakan and other places (Rhodes,2002:1-12).
Stylistically, typologically and metrologically these pieces have been related to the Chandra dynasty of
Arakan. The device of a Bull and Tripartite symbol on the coins of the rulers of Arakan( fig.5) was
directly imitated in the issues bearing the name Harikela. They were prevalent from c. seventh to
twelfth /thirteenth century (Mukherjee, 1975:115-119). These were non dynastic issues. Thus Harikela
as a distinct territorial entity made its presence felt for a pretty long time.

Fluvial Networks as means of Connectivity

With these preliminaries about the geo-political history of Southeastern Bengal, we may now move on
to the fluvial networks that sustained the linkages of this region.
It had been mentioned in the beginning that the distinct sub-regions of Bengal were largely formed by
its hydrography. The name Samataṭa itself signifies a flat coastal area and this is also corroborated by
Xuan Zang (629-645 CE, travels in India). While the Bay of Bengal formed its southern boundary, its
northern boundary was flanked by the river Meghna, and even to the west the Meghna is joined by the
river Padma. Meghna and the Brahmaputra are again linked with each other. The navigability of these
rivers offers excellent riverine networks between the regions through which they flow and connect the
interior to the sea. Fluvial networks of communication can be understood also from the presence of
expressions related to nau, be it a tax or boat parking station, nau-bandhakas or nau-danḍakas (parking
stations for boats and vessels plying in the river) in epigraphic sources of Vaṅga-Samataṭa. To cite an
example we have the Kailan copperplate of Śrīdharaṇarāta (c.CE665-75) where along with reference to
naudaṇḍakas (boat parking stations), we have terms like naupṛthvī, nau-sthira-vegā, nau Śivabhogā in
the context of markers for boundary (Sircar, 1947:221-41) If we take the term naupṛthvī in the literal
sense, it means a land of nau(boats). It is quite possible that in the context of a riverine port, we can
think of a boat scape. Can we then say that naupṛthvī, actually refers to the innumerable number of
boats that were present around Devaparvata, the capital of Samataṭa and a riverine trade centre? The
term kshetra is attached to nau-sthira-vegā. This is a difficult term to explain. But since one of the
boundaries of the donated plot had villa-bhaṅga (bil is a common Bengali word meaning moss covered
with water, watery low lying land) and with it was associated the expression niṣkrāntaka-praviṣṭaka
(according to Chakravarti might be facilities of exit and entry of vessels)( Chakravarti,2002:151) then
perhaps nausthiravega could be a space where the water tended to become stagnant, a sort of watery
low land and this helped the boat to wait there before it could be parked. What is interesting is that in
case of boundary markers this inscription uses terms relating only to boat.

Another interesting example for water bodies being all around a granted area comes from a later
inscription, the Pashchimbhag copper plate of Śrīchandra.(c. 930 CE ). The boundary demarcations on
all four sides of the granted area are recorded: on the east a large embankment (brihatkottāli s̄ imā); on
the south the Mani river (Maninadi s̄ imā: identified with the present Manu river, which emerges from
the hills of Tripura and flows through Maulavibazar of Sylhet); on the west two channels (khaṭaka),
Jujju (identified with Jujnachhera) and Kashṭhaparni and the Vetraghanghi river (= modern Ghunghi
river) and on the north the Kosiyara river (identified with Kusiyara in Sylhet). Thus the boundaries
comprised rivers and streamlets. Ranabir Chakravarti has recently drawn our attention to the very large
area of the naubandhaka, measuring 52 pāṭakas or nearly 2600 bighās of land. This is, therefore, no
ordinary boat-parking facility. Its large size suggests that the area close to the brahmapura had become
an important landmark for inland riverine navigation and communication. According to Chakravarti,
“the presence of a boat-parking station of an impressive size needs to be situated in this context of the
fluvial network of commerce and communication in the Bengal delta” (Chakravarti, 2014:620-621)

In this context of riverine communications, it will be pertinent to bring in Devaparvata here. Our
attention was drawn to the capital of Samataṭa, Devaparvata by Ranabir Chakravarti who suggested
that apart from being an administrative centre it was also a riverine port. (Chakravarti, 2002:151-152).
It was located in the Mainamati ridge near Comilla and archaeological excavations and explorations in
the area have revealed its extraordinary historical as well as archaeological importance.(fig.6) We knew
about Devaparvata from four inscriptions (Husain, 1997:207-231) each of them giving a vivid
description of the landscape of the site. Recently there has been a new addition to the list, the Uḍiśvara
copper-plate. The earliest epigraphic description of Devaparvata is found in the Kailan copper-plate of
Śrīdharaṇarāta (c.AD665-675) (Sircar,1947:221-241) followed by the Uḍiśvara copper-plate of the
same ruler(Islam, 2012:61-72). In the description of the landscape in the Kailan copper plate it is said
that Devaparvata was encircled by the river Kṣirodā (modern Khirnai), both banks of which were
decorated by boats and in which elephants bathed. It is described as sarvotabhadra which means that it
was perhaps approachable on all sides.
(atha-attamātaṅgaśatasukhavigāhyamānovividhatīrthayā naubhir=aparimitābhir=upārachitakulayā
parīkṣitād-abhimatanimnagāminyā Kṣirodayā sarvvatobhadrakād-Devaparvvatāt) The record also
refers to three boat parking stations (naudaṇḍakas) which indicate that it was a riverine port. It was also
the principal political centre of the Rātas. Devaparvata can again be seen in a copper plate of Rājaputra
Balabhaṭṭ̣a, sometimes in the later half of the seventh century CE (Gupta Choudhury, 1979:143). It
speaks of the issuance of the charter from his palace at Kaṭakasila of Devaparvata, a jayaskandhāvāra
which is surrounded by the sacred river Kṣirodā. The next reference to Devaparvata is found in the
Asiatic Society copper plate of Bhavadeva Abhinavamṛgāṅka (c. 765-780 CE) (Sircar, 1951:83-94), a
ruler belonging to the Deva dynasty of Samataṭa. Devaparvata is associated with the river Kṣirodā̄ in
this record too. The city has now gained substantially in stature and glory and its river has become the
most sacred. In the eighth century CE, Devaparvata is explicitly stated to have been the
jayaskandhāvāra which was often coterminous with the royal capital or a major politico-administrative
centre. The last known epigraphic evidence of Devaparvata is furnished by the Paschimbhag copper
plate of Śrichandra (c.925-75CE), dated in his 5th regnal year (Sircar, 1983:92-94). Here too its location
is on the Kṣirodā, on which plied many boats. It is said in the record that Lālambivana (present day
Lalmai, close to Mainamati) was searched by hundreds of boat men for medicinal herbs
(Lālambivanam=atra nāvika-śatair=anviṣṭa). This further speaks of Devaparvata as a riverine port.
(Chakravarti, 2002:167) (Ghosh,2009:352-359). During the time of the Chandras, Devaparvata no
longer figures as a jayaskandhāvāra, though it still was a riverine port. This was due to the more
extensive power base of the Chandras, which necessitated a shift in the capital and to a more viable
geographical location. Here mention should also be made of an a typical topnym
Vaṅgasāgarasambhānḍāriyaka located in Yolāmaṇḍ̣ala, mentioned in the Madanpur copper-plate
charter of Śrichandra. The implication of this toponym was brought to light by Ranabir Chakravarti.
(Chakravarti,2002:144-147) Chakravarti identifies Vaṅgasāgarasambhānḍāriyaka with Sabhar, an
archaeological site, quite close to Dhaka near which the copper-plate has been discovered. The site of
Sabhar is located on the confluence of Vaṃśāvati (Vaṃśāi) and the Dhaleśvari rivers about 17 miles
north-west of Dhaka. It was a riverine port and a centre of Buddhism as large number of clay moulded
tablets with Bodhisattva images, Buddhist stone and bronze sculptures were unearthed.
Vaṅgasāgarasambhānḍāriyaka appears to have functioned as an inland riverine exchange centre,
offered warehousing facilities and provided fluvial communications to the Bay of Bengal
(Vaṅgasāgara). The discovery of the Madanpur copper-plate of Śrichandra from the Sabhar area,
according to Chakravarti, “may therefore strongly suggest that a sambhānḍāriyaka (a trade centre with
warehousing facilities) could very well have existed at Sabhar in the early medieval
times.”(Chakravarti, 2002:148) In an earlier inscription from Vaṅga we have reference to an
administrative centre called Navyāvakāśikā which means a new opening or a channel. Such channels
were mostly navigable and must have offered riverine communication with Vaṅga and the littoral
areas. This linkage also comes alive in the light of a recent study of a hoard discovered from the village
Sandip near Sabhar, consisting of fifteen gold coins (Islam, 2015) of the Samataṭa type, which were
circulating in Samataṭa area after the introduction of the Gupta gold coins in the region which was
already acquainted with Kushāṇa coins. Thus the obverse device of the Gupta coins with the archer
type (depicting the king holding a bow on the obverse) was adopted as the principal obverse device of
the Kushāṇa imitations in circulation in this region (Basu Majumdar 2009: 279-286). What is striking is
that these Samataṭa type of coins were found from Sabhar belonging to the Vaṅga subregion of early
Bengal. Thus these coins were circulating in this region by way of trade. Sabhar perhaps had a
continuous history of being an important trading centre right from around sixth century CE till tenth
century CE and perhaps beyond as till date one can see boats laden with merchandise plying on the
Sitalakshya heading towards Barishal from Dhaka.(fig.7)

Network of interactions of Samataṭa and Harikela

Kāmarūpa located on the ḅank of Brahmaputra and Śrihaṭṭa on the Surma river were included in the
network of interactions of Samataṭa and Harikela both politically and economically. From the
Nidhanpur copper plate of Bhāskara varman (Sharma, 1978:38-54), the ruler of Kāmarūpa we learn
that at least southern Sylhet was within the control of the Varmans of Kāmarūpa from sixth century to
the seventh century CE. This conclusion is on the basis of the location of the land granted which has
been identified with Pañchakhaṇḍa in Sylhet. Kāmarūpa’s proximity to Śrīhaṭṭa gave Bhāskaravarman
access to Samataṭa through southern Sylhet. With this we have the presence of coins in Samataṭa
bearing the legend Śrī-Kumāra identified with Bhāskaravarman.(fig.8) These coins bearing the name of
Śrī-Kumāra were issued in the Samataṭa device and Samataṭa weight standard (ardha satamāna).There
are strong grounds to believe that these heavier and better quality of gold coins (in half satamāna
standard) were peculiar to Samataṭa. Bhāskaravarman did not issue any coin in Kāmarūpa . The seals
used by Lokanātha, Śrīdhāraṇa rata and Maruṇḍanātha, another ruler of the Śrīhaṭṭa region, bear close
similarity and thus indicate some kind of relation between the three within a broad administrative
system. All the three plates are issued from the adhikaraṇa of Kumārāmātya with so-called Gajalakṣmī
(Abhisheka of lakṣmī) emblem, on which these rulers stamped their own seal additionally thus,
suggesting a common pattern.

B.N.Mukherjee earlier suggested while commenting on the Paglatek hoard of Samataṭa coins that the
Parameśvara mentioned in Copperplate of Lokanātha and the Kailan copperplate of Śrīdhāraṇa Rāta
should be the ruler of Kāmarūpa (Mukherjee, 1988:281-286 ). This Parameśvara, the ruler of
Kāmarūpa is likely to be Bhāskaravarman,, though his last date is taken to be 650 or 655 CE, there is
no strong reason for such ascription and could be stretched to at least fifteen more years. Or else it
might so happen that these rulers continued to revere the succeeding Kāmarūpa ruler of the
Śālastambha dynasty for some years even after the demise of Bhāskaravarman, (Ghosh, 2013)

From the Varmans at least a part of Śrihaṭṭa passed on to the hands of the local rulers of Samataṭa, viz.
namely the Nāthas. Sāmanta Lokanātha granted lands to two hundred brāhmaṇas in an aṭavibhukhaṇḍa
in Śrihaṭṭa, obviously with the idea of reclamation of forest lands for expansion of settlement. Next
reference is from the Kalapur plate of Sāmanta MarunḍaNātha (Gupta Chaudhury, 1967:74-80) where
too the grant of a aṭavibhukhanḍa is referred to. Kalapur in southern Sylhet is contiguous to the Comilla
area. After that we have no reference to Śrihaṭṭa in epigraphic sources till the coming of the Chandras
as master of the trans-Meghna. With the expansion of the Candra power the coastal areas of Samataṭa-
Harikela were integrated with northern interiors, located in Śrihaṭṭa and parts of Kāmarūpa , through
the strategic location of Vikramapura in Vaṅga.(Chakravarti, 2011:20).

Two rivers and a seaport are integral to the history and geography of early Assam and Southeastern
Bengal. They are Brahmaputra, Meghna and the port of Samandar. The river Brahmaputra flows
southwest through Assam and south through Bangladesh as the Jamuna. In the vast Ganges Delta, it
merges with the Padma, the main distributary of the Ganges, then the Meghna, before emptying into
the Bahr Harkand, which stood for Bay of Bengal in Arabic sources. The eastern Indian Ocean was
clearly defined as bahr Harkal or the sea of Harikela by the anonymous author of the Hudud al Alam
(c.CE 982). (Minorsky,1937:87). Thus the course of the Brahmaputra suggests its connectivity to river
Meghna and finally to the Bay of Bengal. Along with this Brahmaputra -Meghna system there is
another river which cannot be lost sight of and that is the Surma river. It starts when the Barak
river from Assam divides at the Bangladesh border into the Surma and the Kushiyara rivers. It ends
in the Kishoreganj district above Bhairab Bazar, where the two rivers and their numerous tributaries
and distributaries form a complex network which ultimately flows down as the Meghna river. The
waters from the river finally flow into the Bay of Bengal.

If one takes into account the travels of Ibn Battuta in Bengal in the first half of the fourteenth century,
one notes that he undertook a long journey from Sudkawan or the port of Chittagong to Habang or
Habiganj in Sylhet via Kāmarūpa. He writes that the first city in Bengal that he entered was Sudkawan,
a large town on the coast of the great sea, obviously the Bay of Bengal. Ibn Battuta then recalls that he
set out from Sudkawan for the mountains of Kamaru (Kāmarūpa ) which was a month’s journey from
the Chittagong area. He then journeyed to Habanq, identified with present Habiganj area of Sylhet,
which is traversed by the Blue river, likely to be Meghna. According to him this is used by travellers to
Bengal and Laknawati (Gibbs, 1997: 267-271). Thus Southeastern Bengal appears to have enjoyed
communications and linkages with Kāmarūpa in the Brahmaputra valley. If Ibn Battuta could take this
route in around fourteenth century, it was possible that this was active since earlier times. It should be
mentioned here that Ibn Battuta reached Sudkawan from Maldives by spending forty three nights at sea
(Gibbs, 1997:267). Initially however, he had planned to come to Bengal from Ceylon via Coromandel
coast. The hinterland of this port embraced Kāmarūpa , Arakan and extensive areas in the Ganga
valley. Śrichandra claimed to have conquered Kāmarūpa in the Paschimbhag copper-plate (Sircar,
1967-68:289-304). Verse 12 of the Paschimbhag copper-plate records that Śrichandra’s forces in the
course of their conquest of the Kāmarūpa country entered the woodlands near the Lohitya i.e. the
Brahmaputra where among other things they saw numerous black aloe trees which made the woodland
dark (kāl-aguru-śyāmalaḥ). On the plains of the Lauhitya river leisurely ruminated drowsy yaks.
(Romanthalasa baddhanidrā chāmari-samsevita prāntara Lohitasya vanasthali-parisaraḥ kāl-aguru-
śyāmalaḥ). This indirectly suggests that the raid was also through riverine network as the woodlands
were near Lohita river. We may further postulate that location of the raid was perhaps Haṭappeśvara,
the capital of the Śālasthambha dynasty and identified with present Tezpur. That the banks of
Brahmaputra around Tezpur abounded in black aloes tree is beautifully represented in a verse of the
Tezpur copper-plate of Vanamāla(middle of the 9 th century CE) (Sharma,1978:99). It is said about
Brahmaputra that “this was the river where the waves are rendered fragrant by the clouds formed by the
smokes of black agaru trees burnt by the forest fire occurring in the parks of the city of Haruppesvara”
(tadupavanalapnadāvānaladahyamānakālagurudhūmsambhavāmbudharavṛndasuganjilaugha-
pravāhiṇā). The abundance of aloes wood in Kāmarūpa is also corroborated by the Arab geographers’
accounts which refer to good quality aloes wood from Kāmarūpa and rates it second to the wood of
Multan. They also mention Yak tail as an important commodity of the kingdom of DHM. Al Idrisi
(AD1162) stated that aloe wood was brought to Samandar from Kamrut taking fifteen days by a river
identifiable with the Brahmaputra.( Chakravarti, , 2002 : 165-167.) In one of the notices regarding
Gharu wood, Zhao-ru gua refers to the black aloes wood (Hirth and Rockhill, 1912: 208) and maintains
that it was costlier. Description of forests abounding with black aloes tree in Kāmarūpa finds mention
in inscriptions as mentioned above. Al Idrisi (1162CE) stated that aloe wood was brought to Samandar
from Kamrut taking fifteen days by a river identifiable with the Brahmaputra. From Samandar one
could reach the kingdoms of Southeast Asia. Trade between the kingdom of Kāmarūpa and south
western China through Myanmar is recorded in an Arabic work Al-Masalik wa’l Mamalik. (Sen, 2004:
212). So one might postulate that the black aloes mentioned by Zhao could also be a product of north
east India. The cost of transportation made it costlier. The strong presence of aloes wood as an item of
trade in the Chinese records is clear indicator of the revenue that the coastal polities earned. Needless to
say that its use in the Buddhist rituals enhanced its demand. The prosperity of Kāmarūpa is also linked
with presence of gold in the region from the alluvial deposits of the Brahmaputra. We know that gold
coins were in circulation in Samataṭa. Nicholas Rhodes suggests that Samataṭa’s access to gold may
either have been from Kāmarūpa or from Tibet (Rhodes, 2011: 263-275)

Having a very strong agrarian base and a forest in the vicinity, the major trading items of Samataṭa
must have been agrarian and forest products. Medicinal herbs could also be an important item of trade
as Śrichandra’s inscription talks about sailors looking for medicinal herbs in the forests of Lalmai.
Paucity of source does not allow us to make any inventory of the items of trade. Zhao-ru Gua in the
early thirteenth century refers to cotton from Pong-kie-lo. Pong-kie-lo is generally identified with
Vaṅgāla (referred to in the Chola inscription as territory of the Chandra ruler Govindachandra,
presently in Bangladesh).(Ghosh, 2014:41-54) Here one may also point out that in and around tenth
century CE, Hudud-al Alam also talks about large quantities of good cotton in the Harkand, identifiable
with Harikela region. (Minorsky,1937:87)

Samataṭa’s linkages with regions of Southeast Asia

The best possible cue to have an idea of Samataṭa’s linkage with regions of Southeast Asia comes from
the writings of Xuan Zang (c. 629-645CE, travels in India). Xuan Zang writes that Samataṭa’s frontiers
bordered on the great sea and the very name Samataṭa suggests a flat coastal area. Then he gives a
description of the places in Southeast Asia as having links with Samataṭa or could be easily accessed
from Samataṭa. These are Shi-li-cha-ta-lo( SrikShetra in Myanmar), Kia-mo-land-kia (Kamalāṅka,
identified with Pegu and the Irrawaddy delta in Myanmar), To-lo-po-ti (Dvārāvatī in present Thailand),
I-shung-na-pu-lo (Ishāṇapura, to the east of Dvārāvatī), Mo-ho-chen-po (Mahāchampā in Vietnam) and
Yen-nio-na-chen (identification uncertain, Java according to some). From our numismatic and art
historical sources it is now possible to show the veracity of Xuan Zang’s statement regarding
Southeastern Bengal’s connectivity with regions of Southeast Asia. Samataṭa’s primacy as a
prosperous coastal region with a rich hinterland and significant maritime linkages made it a coveted
area and as suggested by Ranabir Chakravarti experienced the political pull of two pre-eminent eastern
Indian regional powers like Śaśāṅka of Gauda and Bhāskaravarman of Kāmarūpa , none being coastal
authorities.(Chakravarti, 2011:14-17). Samataṭa along with Harikela also experienced continuous
minting of first gold and then silver coins.
We shall begin with the numismatic issue and our case in point is a recently discovered gold coin or
medallion reportedly found from Angkor Borei, Cambodia issued by Īśānavarman I, who reigned in the
early seventh century. (Epinal, 2014:110-111) It has been read (and partly restored) by Arlo Griffiths as
Srīśānavarmma[nah], on the obverse, and īśānapu(ra), on the reverse, meaning, respectively, “of
Īśānavarman” and “Īśānapura.” (Griffiths,2014:56)The king Ishanarvarman is the son of king
Mahendravarman and under his reign, from 611 to 635/637 CE, the power of the pre-Angkorian
Kingdom of Chenla, heir of the Funan empire, stretched from the south of Cambodia to the Mekong
Delta in Vietnam. It is well known from various inscriptions and religious monuments scattered in
Thailand, Cambodia and Laos, that Īśānavarman founded in 615 CE the capital Īśānapura (now
Sambor Prei Kuk), located 35 kilometers north of the city of Kompong Thom. The find spot of the
coin/medallion, Angkor Borei is significant in the political context too. It was the cradle of the Funan
culture which was replaced by Chenla. Thus this medallion could have been issued celebrating the
conquest over Funan. On the obverse we have humped bull resting, facing left, with legs folded under,
except front right leg extended, tail bent with end on ground. Below the bull we have one line Sanskrit
inscription Īśānapu(ra), the city of Iśāna, written in late southern Brāhmī script. The reverse has the
representation of the goddess Śrī/Lakshmī goddess of fortune and prosperity, seated cross-legged on a
lotus flower holding two lotus stems. Above the goddess is a single line inscription in Sanskrit written
in late southern Brāhmī characters, Srīśānavarmma.(fig.9) (Cribb, 2013:9-13)

For our study the significance of this medallion is two- fold. First the device of the coin, bull/seated
Lakshmī is similar to those of the gold coins of Śaśāṅka, the ruler of Gauḍa, circulating in Samataṭa.
Interestingly Joe Cribb opines that the Samataṭa issue of Śaśāṅka is closest in weight and style to the
issue of Īśānavarman. They were struck on a standard of c.5.75grains. It is quite probable that as
Īśānavarman, like Śaśāṅka, was a follower of Śiva, so he preferred to emulate Śaśāṅka’s device, only
exception being the absence of Śiva himself behind Nandi as in Śaśāṅka’s coins.

Secondly the use of the name Īśānapura on the reverse signals the beginning of a tradition of using
place name in coins in Southeast Asia and Southeastern Bengal.
We have seen earlier from Xuan Zang’s accounts that one of the regions with which Samataṭa had links
was Īśānapura. The discovery of this coin betraying adaptation of design from Śaśāṅka’s coins, a
contemporary of Īśānavarman confirms Xuan Zang’s statement. Moreover the Buddhist pilgrim
mentioned that though he could not visit the territories mentioned by him as having linkage with
Samataṭa, yet he could gain knowledge about their customs and manners sitting in Samataṭa. This
statement is indicative of the movement of people among these countries, thereby pointing towards
Samataṭa’s connectivity with the mentioned regions.

The second point of significance assumes a greater dimension when after a close study of the coins
bearing place name in Southeast Asia and Southeastern Bengal, it appears that Īśānavarman’s
medallion/coin is the first one to have the place name written on it. Reference to the name of a kingdom
in silver medallions is also found from Thailand during a little later period. The precise location of
Dvārāvatī was not very sure till the discovery of a number of such silver medals reading on the reverse
‘sridvārāvatīśvarapuṇya’ meaning ‘meritorious deed of the ruler of Dvārāvatī’ at Nakhon Pathom in
1943 (Wicks, 1992:158-159).(fig10). On the obverse it was a cow and calf motif or a flowing vase or
purṇa ghaṭa motif. Other examples have also been found after this from Chainat, Inburi and U Thong
which were a part of the Dvārāvatī cultural zone. The idea of using the name of the kingdom or the
name of a place continued and here mention may be made of the use of the place name Lavapura on
silver coins recovered from a jar near U Thong.(Wicks,1992:158-159) The word Lava was inscribed on
the obverse and pura on the reverse in a late southern Brāhmī script of the seventh or eighth century
CE. Lavapura is identified with the Lopburi cultural period of Thailand.

This practice of using place name in coins has been found in the silver coins from Southeastern Bengal.
Thus we have two significant series of silver coins bearing the name Harikela from the late seventh
century CE. These coins borrowed the device of the Arakanese ruler Dhammavijaya (Rhodes, 2002:1-
12) but Arakanese coins do not have any instance of using place names. Other coins bearing place
name relate to Pilak, Piraka,(fig.11) in Tripura, Vireka, linked with the river Barak, which reaches the
Meghna across the southern border of Sylhet, Pattikera in Samataṭa (Mitchiner,2000:74-75) and
Samataṭa(fig.12) itself.(Mukherjee and Acharjee 2001-2002:81-83)

Thus what transpires from this is that Southeastern Bengal perhaps emulated the practice of using place
names from the Chenla ruler Īśānavarman. There was a continuous process of adaptation and
emulation. While Īśānavarman adapted the device, metal and weight of Śaśāṅka’s coins in Samataṭa, he
introduced the name of his capital in his medallion, while the script used was late southern Brāhmī. In
case of Harikela coins, the device was borrowed from Arakan while the idea of using place name was
emulated from the medallion of Īśānavarman. This was possible due to the continuous movement of
monks and merchants between Southeastern Bengal and kingdoms of Southeast Asia. The knowledge
of place name being used in medallion/coin reached the group of people who minted the Harikela series
of coins. For lack of any strong evidence we are still at a loss to identify the issuer who could be royal
authority or mercantile group. Thus here was a case of adoption from regions of Southeast Asia and not
vice versa. Samataṭa could boast of the riverine port of Devaparvata which was perhaps linked with a
sea port in the Chiittagong area. No coins of the Arakan rulers have been found in the area of the
Chittagong hill tracts which indicate that the chief communication links between the Bengal delta and
central Arakan was by way of the Bay of Bengal and not overland. Definite connections with Arakan
are also proved by use of scripts. The characteristics of the scripts found in eastern India in the seventh-
eighth centuries are found in contemporary inscribed documents from Arakan. Again the letter ‘ha’
used in the name Harikela in the coins is borrowed from Arakan. Taw Sien Ko had shown that the
principal alphabets of Burma, Pyu and Talaing in particular were derived directly from the Kadamba
alphabet of Vanavasi in Karnataka.(Sienko,1921). The inscription of Anandachandra
(Johnston,1944:357-385; Sircar,1958:103-109) found near Akyab, now called the Shittaung inscription
(c.794CE) was written in the Siddhamatrikā script of eastern India and this was due to the close
proximity of Arakan to South –Eastern Bengal.

Nicholas Rhodes draws our attention to a coin of Jishnu Gupta from Nepal who ruled around c.630 CE.
(Rhodes,2011:265-267). On his coin the image of a winged horse in the obverse with a srivatsa symbol
on the reverse could be seen.(fig.13) The artistic treatment of the winged horse appear to be inspired by
a Greek image and Rhodes show that a similar pegasaus image appears on a central Asian coin struck
around c. 600 CE near Bukhara. According to Rhodes the source of the image may have been goods
passing to or from this region of Central Asia. He argues that the most likely source of these images
may have been the silk roundels that appear on silk fabric of this period, which often show winged
animals within a circular border of large pellets. The silk may have been imported from China via the
sea route, up the Bay of Bengal to Samataṭa and thence through Pataliputra through Nepal and further
on to Bukhara and beyond to the Sassanian empire of Iran.
To further elucidate on the route suggested by Rhodes, it can be said that the route could have moved
from China through lower Myanmar, then up the Bay of Bengal to a port in Chittagong which was
linked to riverine ports of Samataṭa and further reached the Ganga valley. Here one is reminded of
Xuanzang’s observations regarding Samataṭa’s linkage with regions of Myanmar mentioned earlier.
The trans-Meghna tracts are actually blend zones between the Ganga delta and main land Southeast
Asia. Therefore the zone was perhaps conducive to linkages with South-east Asia.

Harikela’s connection with regions of lower Myanmar comes in through a completely different kind of
source but more or less of the same period. The text is Mañjuśrimūlakalpa datable to around eighth
century CE. It is interesting to note that in a verse of Mañjuśrimūlakalpa (Vaidya, 1964:508) we have
references to places where the cult of Tārā was prevalent. It reads ‘And then a mighty Yaksha king also
practices Tārā . In Harikela, Karmaraṅga, Kāmarūpa and Kalasa (: sidhyate ca tadā tārā yakṣarāṭ
caiva mahābalaḥ /harikele karmaraṅge ca kāmarūpe kalaśāhvaye // Mmk_53.833 //) What is
noteworthy is that the region of Harikela, identified with the Chittagong region and Samataṭa, identified
with Comilla-Noakhali area, abounds in Buddhist monasteries where we find small bronze images of
Tārā along with other deities. Moreover from the Mainamati copper plate dated CE1220 and issued in
the 17th regnal year of king Ranavāṅkamalla Śrī-Harikāladeva it is learnt that Dhadi-eba, a minister of
the king, donated a piece of land in favour of a vihāra dedicated to Durgottārā ( a form of the Buddhist
goddess Tārā) in the town of Paṭṭikera (Patikara or Paitkara is a paragana near Mainamati). Thus Tārā
was venerated in the coastal region of southeastern Bangladesh. Beautiful clay tablet depicting Śyāma
Tārā has been found from this region.g.) It is very difficult to identify Karmaraṅga. D.C.Sircar places
it in lower Myanmar along with Kalasa as the capital of Karmaraṅga.(Sircar 1976:211-213).
Identification of Kalasa with a place in lower Burma seems plausible and is also agreed upon by Peter
Skilling. If Karmaraṅga is identified with Kamalāṅka as described by Xuan Zang, then it has to be
located in the lower Burma region, close to the sea. Kāmarūpa was of course the present day Assam
and some areas of Northeast India on the bank of the Brahmaputra. It appears from the description of
Mañjuśrimūlakalpa that the regions where Tārā was venerated were close to the sea or river. Thus the
names of Harikela, Kāmarūpa and Kalasa were spelled in the same breath.

Art historical linkages

Dedication of miniature Buddha and Bodhisattva images in Buddhist shrines was a common feature in
South Asia and perhaps was also adopted as a practice in Southeast Asia. It was an important practice
as it entailed merit to the donor. Apart from images, donation of miniature stupas could also be seen.
Donation of minature stupas acquired for the donor the highest religious merit. This we learn from
Xuan Zang when he explains the spiritual importance of these miniature stupas. (Watters, 1961:201)
Numerous miniature metal images have been found from the various monastic sites of Bangladesh
(Mitra,1982). Here we have to categorically mention about Jhewari bronze Buddhas. In 1927, a hoard
of sixty six metal figures was discovered from Jhewari village near Chiittagong. Bronzes from Jhewari
represent Buddha in bhu-sparśa mudrā,(fig.14) Buddha standing in the abhaya-mudrā, seated in
dhyāna or vyākhyāna mudrā, in vajrāsana attended by the Maitreya and Avalokiteśvara, Vajrasattva,
Padmapāni, Mañjusrī,(fig.15) Vasudhārā,(fig.16) etc. (Bhattacharya, 1989:14-17). They range in
height between 3.75 cm and 5.02 cm. Six of the Jhewari images bear undated votive inscriptions which
have been palaeographically assigned between the middle of the ninth and tenth century CE. A few of
the images display features reminiscent of Burmese statuary.(Sengupta, 1989-90:125-130). A group of
bronze images were also excavated at Mainamati , an important Buddhist monastic site in Comilla,
Bangladesh.(Hussain,1997:218). Tārā and Mañjusrī are two most popular iconic types from Mainamati
and Harikela (Fig.17) and it is possible to connect them with bronzes from Peninsular Thailand. To cite
an example, from Hua Khu, Chaiya district a bronze image of eight-armed Tārā has been found which
is housed in the Bangkok National museum. (fig.18) In the opinion of Piriya Krairiksh, the
iconography of this image is similar to that of a figure found in East Bengal.(Krairiksh, 1980:178)
Though Kraikrish has not given further details, it appears that there could be similarity with bronzes
from Mainamati. The iconographic treatment of this image is similar to an image of Mañjusrī from
Mainamati.

There was a strong connection between Javanese bronze and bronzes from Southeastern Bangladesh
during the ninth century CE. According to Pauline Lunsingh Scheurleer, ‘Indo-Javanese bronzes’ did
not originate from Nalanda or any other centers of Bihar as it was earlier believed. They originated
from the Comilla and the Chittagong districts in Bangladesh. According to her these areas were easily
accessible to ships coming from South East Asian countries.( Scheurleer and Klokke, 1988) We are
aware that Java was host to Buddhist pilgrims on their way to China from India. Cultural relations
between India and Java continued throughout the Ancient Indonesian period (5 th to 16th centuries CE).
In the inscriptions and old Javanese literature actual reference to such contacts are found. The
inscription of Kelurak,(778CE) Central Java mentions an Indian guru Kumāraghosha who came from
Gauḍidvīpa. Gauḍidvīpa obviously relates to Gauḍa region of Bengal whose connotation extended to
Bengal and Bihar under the Pālas by the eighth century CE. During the eighth century the port of
Tāmralipta lost its relevance and it is probable that Kumāraghosha went to Java from a port from
Chittagong region. This suggests that movement of monks, teachers, traders from Southeastern Bengal
to Java was not unusual. Therefore there is a possibility that from the eighth century onwards bronzes
from Southeastern Bengal were also imported into Java. At first the Javanese bronze masters copied
them but within a short space of time the indigenous Javanese element predominated. A particular
bronze, bejeweled, four armed Avalokiteśvara in pensive mood, now housed in museum Radyapustaka,
Surakarta is important. It has a solid aureole with flame motifs interspersed on the outer periphery. The
pedestal is a rectangular seat with double-petalled lotus. This bronze image reminds us of the bronze
images from Mainamati or Jhewari in style and treatment. We would like to point out that when
looking for adaptation it is important not to look for the emulation of iconography in its entirety but
elements in sculptures which could have been adapted to an already existing art form. Thus between
Javanese bronzes and Jhewari bronzes we perceive such adaptations in seating posture, aureole and
pedestal. Another pensive bronze Avalokiteśvara with two hands is housed in Rongoworsitto museum,
Semarang, Indonesia. This bronze deity has a strong Pāla influence in the facial expression. These
images of Pensive Avalokiteśvara afford evidence to account for the interaction of styles from different
areas resulting in certain similar stylistic features being borrowed by one centre of production from
another. A comparison between the pedestal and the representation of the lotus seat in the Jhewari
bronzes and the Avalokiteśvara images from Java would point towards interaction between two
ateliers. Thus along with indigenous elements, some commonalty in manufacture and artistic
representation can be perceived.

This commonalty could also relate to the migration of models of famous and potent images. Here we
again take the case of Avalokiteśvara seated in a pensive mood. Of the various forms of this deity, the
‘pensive’ form is unique. Gouriswar Bhattacharya points out that he is called pensive because the
Bodhisattva is seen in a thoughtful gesture but no appropriate nomenclature of the deity from any
literary Buddhist source has been found (Bhattacharya,2002: 125-138) A plaque from the site of
Kutilamura, Mainamati, Bangladesh, datable to c.eighth century CE depicts four-armed
Avalokiteśvara, seated on the principal lotus with attendants of the deity in smaller lotuses.
Avalokiteśvara, is distinguished by Jaṭābhāra and three quarter profile view of the face. (Sengupta,
1989-90: 125). He seems to be in a pensive mood. The sculpture bears the usual Buddhist creed.
(Fig.19) An oval shaped tablet from Nalanda housed in the Asutosh museum bears more or less a
similar iconographical composition. It shows a figure seated in mahārājalīlā on a lotus and has six
arms. His hair is tied in a chignon and a pearl necklace is the only ornament adorning his otherwise
bare torso. His head is inclined to the right and rests against the palm of the uppermost right hand. The
second right hand seems to carry the rosary and the lowest right hand shows varada mudrā. Of the
three left hands only one is visible which is rested on the lotus seat. He is bejeweled. A beaded halo can
be seen around his head. Two lines of Buddhist creed are at the bottom of the tablet. This tablet has
been assigned to 9th century CE. (Fig.20) A similar tablet, little cruder in execution is also housed in the
Ashmolean museum. The idea of the pensive Avalokiteśvara finds expression in a tablet from Yala,
Peninsular Thailand.(Fig.21) His head is slightly tilted to the right and a beaded halo also appears
around his head. His torso is bare too except for a necklace. Another interesting feature of this tablet is
the attempt to make rim with folds, albeit crude, like some of the moulded tablets from Nalanda or
Southeastern Bengal.(fig.22)(Ghosh, 2014:198). Thus the idea of pensive Avalokiteśvara images were
circulating across Asia and this idea gained expression in three mediums of art, stone (Mainamati
sculpture), clay (tablets from Nalanda and Yala) and bronze (sculptures from Java).

Ship Wreck Evidence

In the last two decades several shipwreck sites were excavated in the waters of Southeast Asia and the
materials unearthed offer us a window to the kind of commodities that were being transshipped. For
our topic of discussion it will be worthwhile to mention here about an early to mid-tenth century ship
wreck in the Java sea, the Intan which was excavated in 1997 by Michael Flecker. The cargo is
extremely diverse. There are ceramics, silver ingots, mirrors, and ironware from China; tin ingots and
currency from the Malay Peninsula; fine-paste-ware kendis and bottles from Thailand; and glass and
amphorae from the Middle-East. Indonesian products include bronze ingots, gold coins and jewellery,
aromatic resin, scales sets, and grind-stones. Buddhist figurines, kala-head door knockers, oil lamps,
mirrors, clay tablet moulds, pellet bells, vajra and ghanṭā are all beautifully depicted in Indonesian
bronze. Thus there was an assortment of metal moulds and ritual utensils associated with Mahāyāna
Buddhism.(Flecker, 2002): This shipwreck shows that a locally made vessel traded between Sumatra
and Java, between the powerful empire of Srivijaya and the Javanese state of Mataram. She carried
bronzes cast in Sumatra, yet showing the strong Buddhist and Brahmanical influences of India, Bengal
to be more precise. Since Bengal, like Java, had no tin, seafarers could have previously transported tin
from Kedah to Bengal, where local artisans made the icons, and then shipped these to
Java(Hall,2010:6). The tin was likely to have come from Bangka Island, which was a known region of
the Srivijaya realm, and would have been a logical stopover on the eastern end of the Straits of Melaka
passageway prior to a vessel’s departure for Java.(Manguin, 1982:24-29) In addition, the shipment
included several moulds for local artisans to use to produce bronze and terracotta Buddhist miniature
shrines. The icons produced by these moulds appear in the Borobudur panel relief. These Buddhist
icons according to Hall “were consistent with the international movement of Buddhist clerics and
pilgrims in those times, as Buddhism had been embraced by Java’s kings as a means to transcend
ethnic loyalties and societal institutions”.(Hall, 2010:7). The types of religious artifacts found on the
Intan wreck are even more conspicuous in quantity on another major tenth-century shipwreck,
recovered sixty miles (100 kilometers) off the Java north coast from Cirebon. The Cirebon wreck
carried cargo similar to that found on the Intan vessel, with the exception that it had significantly more
bronze religious artifacts, intended for temple worship, and Buddhist clay moulds, for wider public
consumption, all of which are thought to have been produced in an unknown Sumatra production center
for export to Java. (Hall, 2010:7) Instead of just Bengal we would like to specifically identify this
region as Southeastern Bengal. Is it possible to think that some of these were produced in the
Chittagong region which could boast of a workshop of metal objects? The affluence of a Buddhist
establishment at or near the present Chittagong town for at least three centuries from the ninth to the
end of the eleventh is important enough to draw pilgrim from other places. This Buddhist establishment
could be the Paṇḍita Vihāra mentioned in the Tibetan texts. (Mitra, 1982:26).
 It should be noted that most of these sculptural pieces date between eighth to eleventh century CE both
in Southeastern Bangladesh and Peninsular Thailand. Closely integrated to the making of bronze
sculptures are the raw materials required for making bronze which was an alloy of tin and copper. In
spite of the fact that Bengal did not produce tin we have profusion of bronze sculptures. Thus it had to
establish links, directly or indirectly with tin producing areas. As tin was a trading commodity along
trans-Asiatic routes, it is possible that Southeastern Bengal received its share of tin from the Isthmian
tract. Within the upper part of the Thai- Malay peninsula, the richest tin deposits are located near
Ranong on the west coast and in areas around Nakhon Si Thammarat on the east coast. A later text of
the 14th century, Wang Ta-yuan’s Tao-I Chih-lioh, includes high quality tin among the indigenous
products of Tambralinga.(Hay, 2002:335). There is no doubt that there existed around the isthmus
highly competent sculpture workshops. There could have been migration or movement of sculptors
adept in the making of miniature as well as large sculpture from the Harikela region. Buddhism
flourished in the east coast as port polities were formed. The rulers of these polities patronized the
making and installation of Buddhist images in the locale.

The Buddhist murals of Myanmar too bear testimony to Southeastern Bengal’s interaction with
Myanmar among many other spheres. Claudine Bautze Picron in her study on the murals of Pagan
(Bautze-Picron, 2003:167-168) states that “the source for these murals both from iconographic and
stylistic points of view, might well be sought in north-east India….”. She further states that “while the
style of the murals behind the cult image, or of any Buddha image, finds its closest match in images
from Bihar, the depiction of the Jātakas is more closely related to the pictorial tradition of south-east
Bangladesh.” Strong similarities are seen in the use of colour; beside the dark blue of the background,
red and white dominate, just as in the manuscript illuminations of southeast Bengal. In other words the
stylistic source of these murals lies in the district of Comilla. A case in point is the manuscripts dated in
the reign of Harivarman (c.1073-1127CE), son of Jātavarman of Samataṭa. These manuscripts reflect
an idiom which represents stylistic features of the region also noticed in stone sculptures (plain
background, importance of the outerline delineating the volume) of the Comilla area. Bautze-Picron
rightly states that this area evidently constituted the outer eastern frontier of the “Indian” Buddhist Art
in the eleventh and twelfth centuries and as such was a cultural transitory zone between India and
Myanmar. (Bautze-Picron, 1999:188-189).

Conclusion
In conclusion it may be said that Samataṭa and Harikela’s connected history with regions of Southeast
Asia began as early as the beginning of seventh century CE and continued thereafter. What is important
also to note that Samataṭa in Xuan Zang and Harikela in I jing began to figure prominently in terms of
maritime networks with Southeast Asia even when Tāmralipta was a premier port in the Ganga delta.
From the ninth century onwards, a major port like Samandar (in and around modern Chittagong) was
on the rise. Ranabir Chakravarti (Chakravarti, 2002:166-167) has aptly demonstrated the rise of
Samandar and its links to the feeder ports of the region. Citing reference from Arabic texts, Samandar’s
seaborne contact with Uranshin (Orissa coast), Kanja (Conjeevaram in the northern part of the
Coromandel coast) and Serendib (Sri Lanka) is brought to light by Chakravarti. We have no evidence
of maritime activity from Tamralipta after the eighth century. There after the locus of trade and
shipping moved away from Tamralipta towards Samandar in Chattagram in the south eastern delta.
Saptagram in the western delta emerged much later. Thus there was a hiatus after 8 th century in the
western part of the delta while the port of Samandar in Southeastern Bengal flourished. (Mukherjee,
2011:32) Interaction with Srivijaya during the Pāla period through the maritime route had to be
through the port of Samandar in Chittagong region as by then Tamralipta lost its relevance. This
linkage perhaps helped in the intrusion of Mahāyāna Buddhism in some areas of South East Asia.
Being located in a strategically important region of the international trade, Southeastern Bengal
emerged as an important centre of Buddhism which had a long history of the presence and practice of
Mahāyāna Buddhism. Linkages with Myanmar could be viewed through multiple lenses like
numismatics, paintings, political strategy etc. The port of Samandar/later on Chittagong port was also
used as a transit trade zone for connections between countries across the Bay of Bengal. Thus when in
an account of Bengal given by Marvazi(1120CE) (Minorsky, 1942:147-148) it is stated that ‘in a town
called H.DKIRA(identified with Harikela) business is carried on by means of gold (coins) and cowrie
shells’ but the latter are more in use—’ we know that these shells reached Harikela from Maldives and
was naturally transferred to the hinterlands in the Pāla-Sena domain where cowry was much used as a
parallel currency. Ibn Battuta’s testimony is also evocative of the demand for cowries in Bengal. While
living in Maldives, he procured cowries with some of his jewels and was keen on leaving the island.
But the Wazir was not ready to let him go and persuaded him to stay back. The conversation which
followed between the Wazir and Ibn Battuta is significant. When Ibn Battuta asked the Wazir, “What
shall I do with the Cowries I bought?”; the Wazir replied “Send one of your companies to sell them for
you in Bengal.” The answer is self explanatory and need not be elaborated. (Gibbs, 1997:249).
Samandar/Sudkawan as a port of Bahr Harkand maintained the linkage both with the hinterland
agricultural producers and foreland seafarers reaching out to distant lands.

Acknowledgements

Visits to various sites in Bangladesh were made possible through a financial grant received from the
University of Calcutta under the University Grants Commission’s University of Potential Excellence
project on “Rewriting Regional History: South-Eastern Bengal from c.500CEto c.1300CE”. I am
extremely thankful to the authorities of University of Calcutta for the support. My sincere thanks
go to Prof. Ranabir Chakravarti who generated in me the initial interest in looking at maritime
and riverine networks for understanding linkages. I thank Arindam Mukherjee for encouraging
me to publish this booklet.

Bibliography

Basak, R.G.(Rev.&Trans.) 1969. Rāmacaritam of Sandhyākaranandin, Calcutta:1.5.

Basu Majumdar, Susmita 2009, ‘Typological progression in Numismatic art –A case study of Bengal
gold coinage’, Journal of Bengal Art, Vol.13 & 14: 279-286.
Basu Majumdar,Susmita and Sharmishtha Chatterjee.20014, ‘Cowries in Eastern India : Understanding
Their Role as Ritual Objects and Money’, Journal of Bengal Art, Vol.19:39-56.
Beal, Samuel, 1973(second edition), The Life of Hiuen-Tsiang by the Shaman Hwui Li , New-
Delhi:XXXIX.
Bautze-Picron, Claudine, 1999. ‘Buddhist Painting During the Reign of Harivarman(end of the 11 th
century) in Southeast Bangladesh’, Journal of Bengal Art, Vol.4:159-197.
------------------------------------2003.,The Buddhist Murals of Pagan, Timeless vistas of the cosmos,
Bangkok.

Bhattacharyya, Amitabha, 1977, Historical Geography of Ancient and Early Medieval Bengal,
Calcutta.
Bhattacharya, Asok, K, 1989. Jhewari Bronze Buddhas, A Study in History and Style, Indian Museum,
Calcutta, 1989.
Bhattacharyya, D.C., 1933, ‘The Mainamati Copper-Plate of Ranavankamalla Harikāladeva’, Indian
Historical Quarterly, vol. IX:282-289.
Bhattacharya, Gouriswar, 1993,‘An Inscribed Metal Vase Most Probably From Chittagong,
Bangladesh’, South Asian Archaeology, 1991: 323-338.
------------------------- 2000, ‘A Preliminary Report on the Inscribed Metal Vase from the National
Museum of Bangladesh’, Enamul Haque (ed.), Essays in Buddhist Hindu Jain Iconography and
Epigraphy, Dhaka: 471-487
-------------------------2002, ‘The “Pensive” Bodhisatva From Uddiyana (Swat Valley) To Pattikera
(Mainamati’)—A Unique Example From Magadha ( South Bihar)”, Journal of Bengal Art, 7: 125-138.

-------------------------2014, ‘Pala Influence in Mainamati of Bangladesh’, Journal of Bengal Art, vol.19:

17-27.
Bhattacharya, Swapna 1994, ‘The Ari Cult of Myanmar’, in Gaertner, Uta & Lorenz, Jens eds,
Tradition and Modernity in Myanmar, Muenster: Lit Verlag, pp. 251-271. In the series: Berliner Asien
–Afrika Studien, Humblodt University. (Proceedings of an International Conference held in Berlin on
7-9 May, 1993)

Bhattasali, N.K., 1924, ‘Some image Inscriptions From East Bengal’, Epigraphia Indica, XVII,: 103-
105.

Chakravarti, Ranabir, 2002, ‘Vaṅgasāgara-sambhānḍāriyaka: A Riverine Trade centre of Early

Medieval Bengal’ in Trade And Traders In Early Indian Society, Delhi: 142-159
-----------------------------‘Seafaring in the Bengal Coast: The Early Medieval Scenario’, ibid: 160-183.
---------2011, The Pull Towards The Coast: Politics and Polity In India (C.600-1300CE),Presidential
Address, Ancient India, Indian History Congress, 72nd Session, Patiala: 17-20.
-------------------2014. ‘A Tenth Century Brahmapura at Śrihaṭṭa and Related Issues’ in D.N Jha ed. The
complex heritage of Early India, Essays in Memory of R.S.Sharma, Delhi: 607-625.
Chowdhury, Abdul Momin. 1967, Dynastic History of Bengal, ̣Dacca: 18.
Cribb, Joe, 2013.,‘First Coin of Ancient Khmer Kingdom Discovered’ Numismatique Asiatique, Revue
de la Société de Numismatique Asiatique, Numéro 6Juin:9-13.

Deyell,John S.,2011, ‘Reinterpretation of a Samatata Coin-The First Numismatic Depiction of

Bodhisattva Manjusri’ in Shalilendra Bhandare and Sanjay Garg eds. Felicitas: Essays in Numismatics,
Epigraphy and History in Honour of Joe Cribb, Mumbai:101-106.
Epinal Guillaume, 2014. Cambodge du Funan Au Chenla, 1000 ans d’histoire monetaire, National
Bank of Cambodia, Phnom Phen.
Flecker, Michael, 2002. ‘The Archaeological Excavation of the Tenth-Century Intan Shipwreck,’
BritishArchaeological Reports International Series, 1047 Oxford: Archaoepress
Flemming, Benjamin.J. 2010, ‘New copperplate grant of Śrīcandra (no. 8) from Bangladesh’, Bulletin
of School of Oriental and African Studies, 73: 223-244.
Ghosh, Suchandra. 2006. ‘Character of the Pala Army: Gleanings from Literature and Inscriptions,’
Journal of Asiatic Society of Bangladesh, Humanities Dhaka:206.
--------------------------2011, ‘The Trans Meghna Region: Making of a Sub-Regional Identity,’ Journal
of Ancient Indian History, Vol.XXVII :.220-231.
------------------------- ‘Situating the Local Ruling Houses in the Samataṭa area of the Trans Meghna
region of Bengal (6th to 8th centuries CE): A View from Epigraphy’, Proceedings of the Centenary
Celebrations of the Bangladesh National Museum, Dacca, held in July, 2013 (forthcoming)
------------------------2014,‘Buddhism in Early Medieval South Eastern Bengal from Epigraphy and
Chinese Texts in B.R.Mani et al eds. Pracyabodha, Indian Archaeology and Tradition, Delhi:401-408.
----------------------2014, ‘Locating South Eastern Bengal in the Buddhist network of Bay of Bengal
(c.7th century CE-13th century CE)’ in Proceedings of Indian History Congress, 74th Session, Cuttack:
--------------------2014, ‘South-East Asia and the Eastern Sea Board of India through the Lens of Zhao
Rugua’, in Radhika Seshan ed. Convergence: Rethinking India’s Past, Delhi: 41-54.
------------------2014, ‘Viewing our Shared Past through Buddhist Votive Tablets Across Eastern India,
Bangladesh and Peninsular Thailand’, in Asian Encounters, exploring connected histories, Upinder
Singh and Parul Pandya Dhar (Eds.), Delhi, 2014: 189-201.
Ghosh Suchandra & Sayantani Pal, 2006. ‘Political Geography and Locational Importance of Samatata
– Harikela Region,’ in Jahar Acharjee edited, History-Culture & Coinage of Samatata & Harikela,
Agartala: 78-96.
Gibbs, H.A.R. 1997(second reprint) (transl), Ibn Battuta, Travels in Asia And Africa, New Delhi.
Griffiths, Arlo. 2014, Early indic Inscriptions of Southeast Asia in John Guy ed. Lost Kingdoms Hindu
Buddhist Sculptures of Early Southeast Asia, metropolitan museum of Art, New York:56.

Gupta Choudhury, Kamalakanta., 1979, ‘Two Mainamati Copper Plate Inscriptions of the Khaḍga and
Early Deva Times (7th and 8th centuries A.D.)’, Bangladesh Archaeology, vol. 1, no. 1:141-148.
Gutman, Pamela., 1976, Ancient Arakan with Special Reference to its Cultural History between the 5 th
and 11th Centuries,’ Unpublished Ph.D thesis, Australian National university, Canberra.
Hall, K.R.,2010.‘Indonesia’s Evolving Relationships in the Ninth to Eleventh Centuries: Evidence
From Contemporary Shipwrecks and Epigraphy”, Indonesia, 90, October, :6.-7.
Hirth, Frederic. & W.W.Rockhill ed., and trans.1912. Chau Ju-kua: His work on the Chinese and Arab
Trade in the Twelfth and Thirteenth Centuries, Entitled Chu-fan-chih. 21, St. Petersburg.

Husain, A.B.M., ed. 1997. Mainamati-Devaparvata, The Asiatic Society of Bangladesh, Dhaka:218.

Islam, Shariful., 2012. ‘Uḍiś́vara Copper Plate of Śridhāraṇa Rāta , Journal of Asiatic Society of
Bangladesh, Humanities, Vo.57, no.1June :61-72.
-----------------2014, New Light on the History of Ancient South-East Bengal, Dhaka.

Johnston, F.H., ‘Some Sanskrit Inscriptions of Arakan’ Bulletin of the School of Oriental and African
Studies, vol.II, 1944: 357-385.
Krairiksh, Piriya., 1980, Art in Peninsular Thailand Prior to the Fourteenth Century A.D., The Fine
Arts Department, Bangkok, plate-50:.178.
Lunsingh Scheurleer, Pauline and Marijke J. Klokke, 1988, Divine Bronze, Ancient Indonesian
Bronzes from A.D. 600 to 1600, Rijk Museum, Amsterdam.
Manguin, Pierre-Yves, 1982. ‘The Sumatran Coastline in the Straits of Bangka: New Evidence,’
SPAFA Digest 3,2: 24–29.

Minorsky, V., ed. and trn. 1937, Hudud al ‘Alam, London.

-----------------1942.-trn. and commentary of Sharaf al-Zaman Tahir Marvazi on China, The Turks and
India, London : 147-148; cited by Ranabir Chakravarti in the Appendix of Niharranjan Ray et al. eds. A
Source Book of Indian Civilization, 2000, Calcutta: 639.
Mitchiner, M., 2000. The Land of Water: Coinage and History of Bangladesh and Later Arakan Circa
300 BC to the Present Day, London:
Mitra, Debala., 1982, Bronzes from Bangladesh:a study of Buddhist images from District Chittagong,
Delhi.
Morrison, B.M., 1980, Political Centres and Cultural Regions in Early Bengal, Jaipur-Delhi.
Mukherjee, B.N., 1975, ‘The Original Territory of Harikela’, Bangladesh Lalitkala, vol. 1, no. 2:115-
119.
--------------------------------1982, ‘Commerce and Money in the Central and Western Sectors of Eastern
India’, Indian Museum Bulletin, XVI: 65-83.
-------- 1988, ‘The Paglatek Hoard and The Relation Between Kāmarūpa and Samataṭa’, in
K.K.Dasgupta et al eds. Sraddhānjali, Studies in Ancient Indian History (D.C.Sircar Commemoration
Volume). Delhi:281-286.
-----------1993, Coins and Currency Systems in Post Gupta Bengal, New Delhi.
Mukerjee, B.N & J. Acharjee, 2004, ‘A Coin with the Legend “Samataṭa”’, Numimatic Digest, volumes
25-26 (2001-2002), Nashik:81-83.
Mukherjee Rila, Pelagic Passageways, The Northern Bay of Bengal Before Colonialism, Delhi,
2011:32.
Munro Hay,Stuart, 2002, Nakhon Sri Thammarat, The Archaeology, History and Legends of a
Southern Thai Town,White Lotus, Bangkok :335.
Rhodes, Nicholas., 2002, ‘ The Coinage of Samataṭa, Some Thoughts’, in The Quarterly Review Of
Historical Studies, vol.XLII, Nos 1&2, Kolkata, :1-12.
------------------------2011, ‘Trade in South-East Bengal in the First Millennium CE: The Numismatic
Evidence’ in Rila Mukherjee, ed. Pelagic Passageways, The Northern Bay of Bengal Before
Colonialism, Delhi:.263-275.
Rhodes N.G and S.K Bose.,2003. The Coinage ofAssam, volume I, Pre-Ahom Period, Kolkata-
Guwahati:65-78.
Sengupta, Gautam. 1989-90. ‘Art of South-Eastern Bengal : An Overview’ Journal of Ancient Indian
History. Vol.XIX: parts1-2,: 125.

Sharma, M.M., 1978. ‘Tezpur Copper Plates of Vanamala,’ Inscriptions of Ancient Assam,
Gauhati.:113.
Sircar, D.C., 1947, ‘The Kailan Copper plate Inscription of Śridharana Rāta of Samataṭa’, IHQ,
vol.XXIII:221-41.
----------1951, ‘Copper-Plate Inscription of King Bhavadeva of Devaparvata’, Journal of the
Asiatic Society, Letters, vol. XVII, no. 2:83-94.

---------------------1967-68, ‘Pashchimbhag Copper-Plate of Śrichandra, Year 5.’ Epigraphia Indica,

XXXVII: 289-304.
-------- ----------1973, ‘Paschimbhag Plate of Śricandra, Regnal Year 5’ in Epigraphic Discoveries in
East Pakistan, Calcutta:. 27.
-------------------1982, Śilālekha Tāmraśāsanādir Prasaṅga, (in Bengali), Calcutta.
------------------1982, Pāl-Sen Yuger Vamśānucharit, (in Bengali), Calcutta.
------------------1985, Pāl Pūrva Yuger Vamśānucharit (in Bengali), Calcutta.
-------------------1993, Select Inscriptions Bearing on Indian History and Civilization, vol. 1, Delhi.
Takakusu, J., (tr.), 1896, A Record of the Buddhist Religion as practiced in India and the Malay
Archipelago (A.D.671-695) by I-tsing, Oxford, : xxxiii, xlvi, liii and 44
Taw Sien Ko, 1921, List of Inscriptions found in Burma, Part I, Rangoon, p.iv.
Watters, T. 1961, On Yuan Chwang’s Travels, New Delhi.
Wicks, Robert.1992, Money, Markets and Trade in Early Southeast Asia,New York.
Vaidya,P.L.ed.1964, AryāMañjuśrimūlakalpa, (Mahayanasutrasamgraha, part II, Buddhist Sanskrit
Texts, no.18), Darbhanga :508.
The other three subregions were Puṇḍravardhana (initially north Bengal, but gradually it encompassed a large
area of early Bengal), Rāḍha (mainly to the west of the river Bhagirathi), Vaṅga (central deltaic Bengal, covering
the present Dhaka-Vikrampur-Faridpur area, Bangladesh) .
This ruler is in all probability Bhāskaravarman of Kāmarūpa.
We have coins with Pattikera written on it.
Chakravarti gives a detailed analyses of both the terms Vaṅgasāgara and sambhānḍāriyaka and shows that the
eastern sea or Bay of Bengal was called Vaṅgasāgara in the third quarter of the tenth century CE. He also
highlighted the importance of the site of Sabhar identifiable with Vaṅgasāgarasambhānḍāriyaka .
I am extremely thankful to Dr. Shariful Islam of the Bangladesh National Museum for allowing me to refer to
his unpublished paper entitled ‘A Hoard of Unpublished Post-Gupta Gold Coins from Sabhar’ presented during
the 11th International Congress on Bengal Art held between 7 th -10th February, 2015. This hoard was discovered
in 2002 and is presently in the possession of the Bangladesh National Museum.

It is however to be noted that though during the heyday of Devaparvata (7 th century to 9th century CE) as a
riverine port we do not have much information about Sabhar, it must have been to some extent active to grain
primacy again during the reign of the Chandras.
The Zhu Fan zhi of Zhao Ru-gua written in 1225/26 CE is an extremely important document for history of
maritime trade. Zhao Ru-gua, a customs official, looks at the maritime scenario from the perspective of Song
China. The Song Government perceived the sea as an active space of activity both for its exports and imports.
Rudravarman, the last king of Funan, declared Angkor Borei his capital in 514 CE. Chinese dynastic
descriptions of the Funan state disappear in the mid sixth century CE and Tang dynastic records
reports the conquest of Funan by Chenla in the early seventh century CE and archaeological evidence
from Oc eo suggest that the Funan port town was abandoned by the mid seventh century CE. Please
See Miriam Stark, “ From Funan to Angkor: Collapse and Regeneration in Ancient Cambodia” in
After Collapse, The Regeneration of Complex Societies, edited by Glen. M. Schwartz and John. J.
Nichols, 2006, University of Arizona Press, Tuscon, pp.144-167.
Originally the Lakshmi device was taken to be the obverse and the bull as the reverse by scholars who have
published the coin but Joe Cribb rightly showed that the Bull device should be the obverse device following the
coin of Śaśāṅka which was the prototype for this coin/medallion.
For a general history of Arakan please see Pamela Gutman., 1976, Ancient Arakan with Special
Reference to its Cultural History between the 5 th and 11th Centuries,’ Unpublished Ph.D thesis,
Australian National university, Canberra.

Personal communication from Peter Skilling. I am grateful to him for drawing my attention to this essay of
D.C.Sircar.
It is interesting to note that a gold coin from Samatata bears the depiction of Mañjusrī on the obverse. See John,
S. Deyell, ‘Reinterpretation of a Samatata Coin-The First Numismatic Depiction of Bodhisattva Manjusri’ in
Shalilendra Bhandare and Sanjay Garg eds. Felicitas: Essays in Numismatics, Epigraphy and History in Honour
of Joe Cribb, Mumbai:101-106.

Bin Yang suggests that the origin of kauri in Yunnan was Bengal, in other words Bengal was the source of
kauris in Yunnan. From the records of the Chinese travellers he has been able to trace the route from Bengal to
Yunnan through mainland Southeast Asia. Thus Bin Yang correctly identifies an economic network centering
kauri which was used as a parallel currency in Bengal. For a general overview of Cowries in Eastern India and
their role see Susmita Basu Majumdar and Sharmishtha Chatterjee, ‘Cowries in Eastern India : Understanding
Their Role as Ritual Objects and Money’, Journal of Bengal Art, Vol.19, 2014:39-56.

PAGE \* MERGEFORMAT 38